Legacy effects in prairie restoration

In the past dozen or so years ecologists have begun to realize that the future of a prairie remnant is strongly influenced by its past history, and these “legacy effects” are important in restoration work.

I stumbled on these concepts of historical ecology while working on the restoration of Black Earth Rettenmund Prairie, a State Natural Area. Before it became a Prairie Enthusiasts’ site, this outstanding natural area was owned by the Nature Conservancy (TNC), and TNC had extensive files on its history. Also, there is an extensive sequence of air photos to follow the changes through the years.

In 1937, when the first air photo was taken by the U.S. Soil Conservation Service, the prairie was almost completely devoid of woody vegetation. During the four decades between the early period and 1980 there was a gradual increase in wooded area, and the rate accelerated between 1980 and 1987 (see graph below). The early shrub/tree development took place in the south part of the prairie, which is adjacent to a fairly extensive undisturbed woodlot. By 1949, the woody patches had increased, and some of the trees in the South unit were now quite large. (The stumps from some of these trees still remain today.)

Although only approximate, the rate of increase in woody vegetation has some resemblance to a growth curve. There is an extended lag phase in the early years, but beginning in the mid 1970s there seems to be a “tipping point” after which the site seems to “explode” with woody vegetation. (The terms “tipping point” and “explode” were first suggested to me by Randy Hoffman. According to him, left unchecked other prairie remnants in Wisconsin show this same phenomenon. The graph shows that at the peak, almost half of the site was wooded.

Upon acquisition in 1986, this preserve became a major project of the DNR and TNC. Tree and shrub removal and frequent prescribed burns by the DNR and TNC were used in attempts to restore the prairie.

Although these efforts led to substantial success, the legacy of the woody vegetation remains, considerably complicating restoration efforts. New shoots still appeared in former aspen areas. Sumac and gray dogwood were particularly troublesome, but brambles and grape were also a problem. One former wooded area had been invaded by willow and hazel, which have proved difficult to eradicate.

I gave a presentation on my work at the North American Prairie Conference in Winnipeg, Manitoba, in 2012.  My paper has now been published and can be downloaded as a PDF using this URL: This Blog Post is a brief summary.

 

A view looking SW from Fesenfeld Road, showing the extensive clone of aspen on the top and sides of the north unit. Note also the large clone of sumac on the lower slope (above the cropped field). 1986 photo from TNC archives taken by a real estate assessor.

A view of the “Saddle” area of Black Earth Rettenmund Prairie, probably taken in the late 1970s. In addition to the substantial brush, the edge of a large aspen clone can be seen in the lower right corner. Photo kindly provided by Cliff Germain, retired head of the Bureau of Endangered Resources, Wisconsin Department of Natural Resources.

Woody vegetation for different years as measured on air photos by GIS. (a) Thumbnails for 1937, 1987, and 2010. (b) Change in woody areas with time over the 73-year period.

The present work suggests some principles of importance for restoration ecology.

·       You can’t let things get out of hand. During the early stages, shrub establishment may seem benign and not worth worrying about. This is obviously wrong. At first, shrubs and trees are easy to remove. However, once the explosive phase of growth has set in, not only is the biomass needed to be removed much larger, but the whole population is growing at a vastly increased (quasi exponential) rate.

·       The importance of frequent fires. Fire is now a widely accepted component of prairie and savanna management systems. However, there is a tendency to assume that fire on a three- to four-year cycle may be sufficient. This is a mistake. Beginning at the time of TNC acquisition, burns have been conducted on the various units on the average every second year. Despite this frequency, shrub growth remains a problem. Sumac, prairie willow, brambles (Rubus spp.), and gray dogwood in particular continued to thrive. Aspen shoots continue to appear in former aspen areas. In those areas where the legacy of shrub growth remains, annual fire should be considered. The historical GIS data will help locate the areas where annual fires should be carried out.

·       However, burns alone are not enough. At Black Earth Rettenmund Prairie, with its biennial burn cycle, sumac, gray dogwood, prairie willow, etc. have remained or have developed into problems. There are strong reasons to believe that all clonal shrubs, even if they are native, should be eradicated by a multi-year program of herbicide use. If the prairie is burned in alternate years in the spring, then the shrubs should be treated with herbicide in the late fall of the second year after the burn (that is, before the upcoming spring burn).

·       Eradication of clonal shrubs is essential, even if native, because annual stewardship cannot be guaranteed to continue into the indefinite future.

·       Legacy effects must be taken into consideration when planning management strategies. Areas that were woody in 1986, when restoration began, continue to inhibit the established of true prairie species, especially graminoids. These areas not only need tree and brush removal, but planting with seed collected elsewhere on the site.

·       Significance of herbicide use. The prejudices against use of herbicides remain. The first seven years of restoration work at Black Earth Rettenmund Prairie were wasted by trying to control shrubs without the use of herbicides. (The State Natural Area Board had refused permission to use herbicides.) Although trees can often be controlled without herbicide (by girdling), shrubs, with their paramount ability to resprout, cannot be controlled without herbicide.

·       Seed collecting and overseeding are essential. A site that has been wooded for an extended period of time almost certainly has lost most or all of its warm season grasses, as well as some of the more light-demanding forbs. The seed bank for such a site is uncertain, and probably highly variable. Without overseeding, restoration of such a site may take years. Even if there is a good seed source near by, spontaneous overseeding will be random and, depending on the quality of the seed, the success will be highly variable. Even 25 years after the initial restoration work, formerly wooded areas at Black Earth Rettenmund Prairie still remained impoverished, even though they have been reseeded with native grasses and forbs. Although each restoration project must be analyzed individually, if the site was wooded it can safely be assumed that overseeding is advisable. In fact, why not plant? Unless this is a research site, it seems better to plant than to wait and see what happens. Overseeding is easy and relatively inexpensive, and the sooner it is begun the better. The only reason not to overseed may be if a local seed source is not available.

For some reason, Black Earth Rettenmund Prairie did not become infested with red cedar, as have all the other hillsides in the area. This was fortunate, because the shade imparted by red cedar is much worse than that of aspen or shrubs. Cedar-infested sites quickly lose all of their species diversity. In one study at Konza Prairie, herbaceous cover decreased over 99%. (See Briggs, John M., Knapp, Alan K. and Brock, Brent L. 2002. Expansion of woody plants in tallgrass prairie: a fifteen-year study of fire and fire-grazing interactions. American Midland Naturalist 147: 287-294.)

(I am grateful to Wayne and Sharon Gaskill, stewards for TNC, for sharing with me their detailed stewardship files.)